Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=cgsj20

Download by: [Bilkent University] Date: 27 November 2017, At: 22:11

Global Society

ISSN: 1360-0826 (Print) 1469-798X (Online) Journal homepage: http://www.tandfonline.com/loi/cgsj20

Leadership Analysis at a “Great Distance”: Usingthe Operational Code Construct to AnalyseIslamist Leaders

Özgür Özdamar

To cite this article: Özgür Özdamar (2017) Leadership Analysis at a “Great Distance”: Using theOperational Code Construct to Analyse Islamist Leaders, Global Society, 31:2, 167-198, DOI:10.1080/13600826.2016.1269056

To link to this article: https://doi.org/10.1080/13600826.2016.1269056

Published online: 09 May 2017.

Submit your article to this journal

Article views: 100

View related articles

View Crossmark data

Citing articles: 1 View citing articles

Leadership Analysis at a “Great Distance”: Using theOperational Code Construct to Analyse Islamist Leaders

ÖZGÜR ÖZDAMAR

Operational code analysis was originally designed to analyse the Soviet Politburomembers’ political strategies and was codified by George and Walker, Schafer andYoung. The approach has mostly been used to analyse individual world leaders, ratherthan specific political groups. In this article, I use the operational code construct toanalyse political Islamists’ belief systems. I discuss issues regarding using the operationalcode construct in a non-Western political and cultural setting. The first difficulty of usingthe operational code analysis to analyse political Islamists’ strategies is the diversity of themovement, which led to a certain “level of analysis” problem. Second, whether to usecomputerised analysis or traditional hand coding was an important decision. Third,some procedures of the operational code research agenda, such as coding only transitiveverbs, at times seemed to be in conflict with Turkish and Arabic sentence structures andverbs. Lastly, since the “Western” foreign policy analysis (FPA) theories are not fre-quently applied in the “global south” cases, I did not have a large body of literature touse while conducting my research. This article discusses the “boundedness” of the oper-ational code analysis in analysis of non-Western individual leaders and political move-ments by discussion of these issues and others.

Introduction

Operational code analysis1 was originally designed by Nathan Leites to analyseSoviet Politburo members’ political strategies, and was codified by AlexanderGeorge2 and Walker, Schafer and Young.3 In his monograph The Operational Codeof the Politburo, Leites deals with issues such as determining the Bolsheviks’ rulesand exploring the effects of Lenin’s and Stalin’s writings on that large politicalgroup. Leites’ findings led him to make some generalisations and predictionsabout Politburo members’ political moves. Operational code analysis wasfurther developed, first by Alexander George4 and then by Walker, Schafer and

1. Nathan Leites, The Operational Code of the Politburo (New York: McGraw-Hill, 1951); Nathan Leites,A Study of Bolshevism (New York: Free Press, 1953).2. Alexander L. George, “The Operational Code: A Neglected Approach to the Study of Political

Leaders and Decision Making”, International Studies Quarterly, Vol. 23, No. 2 (1969), pp. 190–222.3. Stephen G. Walker, Mark Schafer and Michael D. Young, “Systematic Procedures for Operational

Code Analysis: Measuring and Modeling Jimmy Carter’s Operational Code”, International Studies Quar-terly, Vol. 42, No. 1 (1998), pp. 175–190.4. George, “The Operational Code”, op. cit.

Global Society, 2017Vol. 31, No. 2, 167–198, https://doi.org/10.1080/13600826.2016.1269056

© 2017 University of Kent

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

Young.5 Since George’s call to revitalise the research agenda, it has mostly beenused to analyse world leaders rather than political groups.6 The operational codeconstruct has become an important research agenda for cognitively orienteddecision-making researchers within the foreign policy analysis (FPA) literature.In this article, I use the operational code construct to analyse representativeleaders of political groups that appear and operate in a region far from NorthAmerica where this literature was born. That is, I use operational code analysisto analyse the belief systems of political Islamists in the Middle East and NorthAfrica (MENA) and discuss difficulties associated with this endeavour.Political Islam became an influential movement in the second half of the 20th

century in the MENA. Built on classic Islamic sources, such as the Quran and theSunna, Muslim history and modern political ideologies (such as socialism, nation-alism and anti-capitalism) have also shaped the ideology. With the end of the ColdWar, the political movement reasserted itself in democratic politics (such as inTurkey with the Welfare Party [RP] and the Justice and Development Party[AKP]) and also via illegal organisations (such as the Muslim Brotherhood inEgypt) and armed groups (such as Hezbollah) in the region and the rest of theworld (e.g. the Pan Malaysian Islamic Party in Malaysia or the Prosperous JusticeParty in Indonesia). All these movements, however, have been influenced bysimilar intellectual sources. In addition to the classic sources, followers of politicalIslam read Muslim authors such as Seyd Qutb, Ali Shariati, Mohammed al-Ghazaliand Said Nursi, as well as politicians like Ruhollah Khomeini.While using insights from the operational code literature to analyse belief

systems of Islamists in the MENA, I encountered some issues regarding thetheory, methodology and general applicability and boundedness of operationalcode analysis in and to different political and cultural settings. Although oper-ational code analysis was originally developed to analyse a non-American group(the Bolshevik party in the USSR), modern methods tend to be technical and auto-mated, and therefore problematic when applied to different political and culturalsettings. The first difficulty in using operational code analysis to analyse politicalIslamists is the diversity of political Islam; it is neither cohesive like the Politburo,nor are there many individual Islamist leader examples from the same country. Forthese reasons, I encountered a certain “level of analysis” problem during myresearch; I had to compare different leaders operating in different polities. Thesecond issue involves language. Modern quantitative operational code analysis isbased on English texts and automated systems. When using an automatedsystem, researchers gain in terms of reliability but lose considerably in terms of val-idity. Although hand coding also reveals mixed results, I decided to hand code thedata due to the limitations of the automated system. Third, some procedures of theoperational code research agenda, such as coding only transitive verbs, at timesconflicted with Turkish and Arabic sentence structures and verbs. Finally, sinceWestern FPA theories are not frequently applied in “global south” cases, I didnot have a large body of literature to resort to while conducting my research.This article discusses the boundedness of the operational code construct by ana-

lysing non-Western individual leaders and political movements. It begins withbrief introductions to political Islam and operational code analysis to inform the

5. Walker, Schafer and Young, op. cit.6. George, “The Operational Code”, op. cit.

168 Leadership Analysis at a “Great Distance”

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

reader about the subjects at hand. A summary of my research on the operationalcodes of original and neo-Islamists follows. Finally, I discuss theoretical, methodo-logical and region-specific issues related to applying the operational code constructto Islamist leaders. I devote the final section to the debate on the boundedness ofNorth American FPA theories by using insights from this research.

Political Islam and Middle Eastern Politics

Political Islam reappeared as a powerful political force in theMENAduring the secondhalf of the 20th century. Although it is certainly not a “monolithic phenomenon”,7

partly as a reaction to colonisers and the superpower penetration of the region, andpartly as a reaction to the failure of domestic secular nationalist movements toachieve political and economic development, political Islam regained strengthespecially after the late 1960s.8 Islamist movements achieved success in Iran,Lebanon, Palestine and Libya, were harshly suppressed in Algeria and Egypt, andhave produced coalition partners and election-winning catch-all parties in Turkey.

For scholars of Islam and theMiddle East, there has never been a clear distinctionbetween religion and the state in the classical Islamic tradition. This was reflected inthe Muslim empires’ foreign policy, conception of war and peace, and internationallaw.9 Islamic concepts were—to some extent—taken into consideration whenmaking decisions in these foreign policy-related issues. Yet, with the Iranian revolu-tion and its aftermath, when researchers attempted to understand the influence ofIslam on foreign policy decisions, they did not succeed completely due to the under-development of political science methodologies to evaluate Islamic influences onpolicy.10 There have been attempts, however, from both the Islamic world11 analys-ing foreign affairs from a purist Islamic perspective and from Western academics12

using mainstream Western geopolitical theories. Nevertheless, literature on Islam’sinfluence on foreign policy in predominantly Muslim countries remains in itsinfancy. After 9/11, the US-ledwars in Afghanistan and Iraq, and the Arab uprisingssince 2011, there has been a resurgent interest in the subject.

A few studies that deal with non-Western leadership by using the operationalcode construct have been conducted in the last two decades, although they donot focus on political Islam per se. Crichlow uses the operational code constructto analyse whether beliefs have an impact on Israeli foreign policy decisions.13

The results showed fluctuating beliefs due to changing dynamics in the MiddleEast and in the world. Dyson has used the operational code construct to analyse

7. Muhammad Ayoob, “The Future of Political Islam: The Importance of External Variables”, Inter-national Affairs (Royal Institute of International Affairs 1944-), Vol. 81, No. 5 (2005), pp. 951–961. Retrievedfrom http://www.jstor.org/stable/3569069.8. Anthony Black, The History of Islamic Political Thought: From the Prophet to the Present (New York:

Routledge, 2001).9. Bernard Lewis, The Political Language of Islam (Chicago, IL: University of Chicago Press, 1991);

Bernard Lewis, The Middle East: The Brief History of the Last 2,000 Years (New York: Scribner, 1995).10. Adeed Dawisha, Islam in Foreign Policy (Cambridge: Cambridge University Press, 1983).11. Abdul Hamid Abu Suleyman, The Islamic Theory of International Relations: New Directions for Islamic

Methodology and Thought (Herndon, VA: International Institute of Islamic Thought, 1987).12. Graham E. Fuller and Ian O. Lesser, A Sense of Siege: The Geopolitics of Islam and the West (Boulder,

CO: Westview, 1995).13. Scott Crichlow, “Idealism or Pragmatism? An Operational Code Analysis of Yitzhak Rabin and

Shimon Peres”, Political Psychology, Vol. 19, No. 4 (1998), pp. 683–706.

Özgür Özdamar 169

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

then-new Russian leader Vladimir Putin.14 Feng analyses Chinese leaders vis-à-visthe straits crisis.15 Malici and Malici have focused on two communist leaders—Fidel Castro and Kim Il Sung—to analyse whether the end of Cold War madeany change to their beliefs about the political universe and their tactics.16 Theirresults show there was only a modest change despite the systemic shifts. Maliciand Buckner show that seemingly “rogue” leaders are not as hostile as convention-al wisdom suggests and that they have tactics designed to avert escalation andcrisis.17 Devlen uses operational code analysis to analyse the Iranian nuclear pro-gramme row.18 Devlen’s analysis suggests that although Iranians are keen to nego-tiate with the US, their general view of the US is very conflictual, which preventedengagement in constructive diplomacy. Similarly, O’Reilly suggests that beliefshave a direct impact on even the most strategic decisions such as nuclear prolifer-ation.19 His study focuses on South African and Indian decision-making in regardto building nuclear weapons, and supports the notion that actors’ perception of theinternational system is more important than abstract notions such as the securitydilemma when making the most strategic decisions. Even though there havebeen some examples of operational code analysis in non-Western cases, applicationof the construct to Muslim leaders has been very rare, with Malici and Buckner oneof the few exceptions.20

Despite its importance in the MENA in the last half-century, political Islam is alsounderstudied. Most studies that analyse these movements suffer either from politicaland cultural biases or a lack of understanding about the region. In the FPA literature,few studies focus on political Islam as a distinct political ideology with specificforeign policy preferences. This research aims to fill the gap.21 I believe that themicro-level analysis of FPA offers some of the most powerful methodologies toapproach the question of political Islam’s influence on foreign policy. More specifi-cally, FPA allows us to see this linkage through the analysis of leader beliefs. Accord-ingly, this study will analyse the belief systems of Islamist leaders. First, I focus onthree Islamist leaders of the previous generation: Necmettin Erbakan of Turkey,ImamKhomeini of Iran, andMoammerQaddafi of Libya. I then turn to neo-Islamists

14. Stephen Benedict Dyson, “Drawing Policy Implications from the ‘Operational Code’ of a ‘New’Political Actor: Russian President Vladimir Putin”, Policy Sciences, Vol. 34, No. 3/4 (2001), pp. 329–346,available: <http://www.jstor.org/stable/4532539>.15. Huiyun Feng, “Crisis Deferred: An Operational Code Analysis of Chinese Leaders across the

Strait”, in Mark Schafer and Stephen Walker (eds.), Beliefs and Leadership in World Politics: Methods andApplications of Operational Code Analysis (New York: Palgrave Macmillan, 2006), pp. 151–170.16. AkanMalici and JohnnaMalici, “The Operational Codes of Fidel Castro and Kim Il Sung: The Last

Cold Warriors?”, Political Psychology, Vol. 26, No. 3 (2005), pp. 387–412, available: <http://www.jstor.org/stable/3792603>.17. AkanMalici and Allison L. Buckner, “Empathizing with Rogue Leaders: Mahmoud Ahmadinejad

and Bashar al-Asad”, Journal of Peace Research, Vol. 45, No. 6 (2008), pp. 783–800.18. Balkan Devlen, “Dealing or Dueling with the United States? Explaining and Predicting Iranian

Behavior during the Nuclear Crisis”, International Studies Review, Vol. 12, No. 1 (2010), pp. 53–68, avail-able: <http://www.jstor.org/stable/40730709>.19. K.P. O’Reilly, “Leaders’ Perceptions and Nuclear Proliferation: A Political Psychology Approach to

Proliferation”, Political Psychology, Vol. 33, No. 6 (2012), pp. 767–789, available: <http://www.jstor.org/stable/23324191>.20. Malici and Buckner, op. cit.21. This research about the belief systems of political Islamists was presented at the annual meetings

of the International Studies Association in 2011 and 2012. It is still in progress and not published else-where. In this article, a summary of the research process and broad conclusions are presented.

170 Leadership Analysis at a “Great Distance”

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

and perform an operational code analysis of Prime Minister Tayyip Erdoğan ofTurkey, Mahmoud Ahmadinejad of Iran, and Khalid Meshal of Gaza. To analysethese leaders’ belief systems about foreign policy, I use the operational code approach.

Operational Code: Beliefs and Strategies

Operational code analysis focuses on the beliefs of political leaders as causal mech-anisms in explaining foreign policy decisions. It was originally developed byLeites22,23 to analyse the decision-making style of the members of the Soviet Polit-buro and later refined by George,24,25 Holsti26 and Walker27,28. According to theoperational code construct, a leader’s cognitive schema or belief system has twocomponents. The first set contains five philosophical beliefs about the leader’s politi-cal universe and the nature of the “other” he or she faces in this environment. Thesecond set holds five instrumental beliefs that represent the image of “self” in thispolitical universe and the best strategies and tactics a leader could employ toachieve his or her ends.29 Taken together, the two sets explain decision-makers’ ten-dencies on foreign policy.30

A leader’s philosophical and instrumental beliefs are the answers to a set of ques-tions developed by George.31 Regarding philosophical beliefs, the questions are:

P-1. What is the “essential” nature of political life? Is the political universeessentially one of harmony or conflict? What is the fundamental characterof one’s political opponents?P-2. What are the prospects for the eventual realisation of one’s fundamen-tal values and aspirations? Can one be optimistic, or must one be pessi-mistic on this score, and in what respects the one and/or the other?P-3. Is the political future predictable? In what sense and to what extent?P-4. Howmuch “control” or “mastery” do self and other have over histori-cal development? What is self and other’s role in “moving” and “shaping”history in the desired direction?P-5. What is the role of “chance” in human affairs and in historicaldevelopment?

Concerning instrumental beliefs, the questions are:

I-1. What is the best approach for selecting goals or objectives for politicalaction?

22. Leites, “The Operational Code”, op. cit.23. Leites, “A Study of Bolshevism”, op. cit.24. George, “The Operational Code”, op. cit.25. Alexander L. George, “The Causal Nexus between Beliefs and Behavior”, in Lawrence S. Falk-

owski (ed.), Psychological Models in International Politics (Boulder, CO: Westview, 1979), pp. 95–124.26. Ole Holsti, “The Operational Code as an Approach to the Analysis of Belief Systems”, Final Report

to the National Science Foundation (1977), Grant No. SOC 75-15368.27. Walker G. Stephen, "The Motivational Foundations of Political Belief Systems: A Re-analysis of the

Operational Code Construct", International Studies Quarterly, Vol. 27, No. 2 (1983), pp. 179–201.28. Stephen G. Walker, “The Evolution of Operational Code Analysis”, Political Psychology, Vol. 11,

No. 2 (1990), pp. 403–418.29. George, “The Causal Nexus”, op. cit.; Walker, “The Evolution of”, op. cit.30. Schafer and Walker, op. cit.31. George, “The Operational Code”, op. cit.; George, “The Causal Nexus”, op. cit.

Özgür Özdamar 171

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

I-2. How are the goals of action pursued most effectively?I-3. How are the risks of political action calculated, controlled and accepted?I-4. What is the best “timing” of action to advance one’s interests?I-5. What is the utility and role of different means for advancing one’sinterests?

P-1, I-1 and P-4 are considered master beliefs; they are key in determining theanswers to the other questions. The central assumption of operational code analysisis that individual leaders matter in shaping states’ foreign policies and that thebeliefs a leader has might act as causal mechanisms in understanding why he orshe chooses a certain foreign policy decision. Conversely, mainstream internationalrelations theories (such as neorealism) rule out diversities in leaders’ beliefs, theireffects on foreign policy behaviour, and leaders’misperceptions of their states’ pos-itions as causes of leaders’ effects on the international system.32

Trying to answer George’s questions about philosophical and instrumentalbeliefs,33 Holsti developed six types of operational codes (A, B, C, D, E, F),34

which Walker later reduced to four (A, B, C, DEF), op. cit.35 This typology isbased on the nature (temporary versus permanent) and the source (individual/society/international system) of conflict in the political world, deduced from theanswers to P-1, I-1 and P-4.36 In the revised typology, pessimists (DEF types),who see conflict as permanent, are defined as a single group regardless of thesource of that conflict. Walker argues that such a move is justified because thereare no discernible differences in the remaining philosophical and instrumentalbeliefs of DEF types, thus they can be seen as a single ideal type.37 Optimists,however, differ in how they see the source of conflict: misperceptions by the indi-viduals (A), societal institutions (B) or the anarchic nature of the internationalsystem (C). Those differences have an impact on the rest of the leader’s philosophi-cal and instrumental beliefs.38 Table 1 represents this typology.The four quadrants that represent six different leadership types also propose a

specific preference ordering for each type. Modern operational code analysis sub-scribes to a via media position in the rational choice–cognitive theories debate andincorporates strengths of both. The main argument of operational code analysisis that preferences are not fixed for agents as rational choice theory assumes. Byfocusing on specific belief systems of leaders, researchers can infer actors’ prefer-ences in a more robust fashion. The leadership types and corresponding preferenceorderings are in line with the main assumptions of operational code analysis andgame theory research. Type A and C leaders, who have a more friendly orientation,prefer to settle their differences with political opponents rather than resorting todomination, deadlock or submission. Type B and DEF leaders prefer dominationover other preferences. These preference orders are used to construct various 2×2

32. Schafer and Walker, op. cit., p. 8.33. George, “The Operational Code”, op. cit.34. Ole Holsti, “The Operational Code”, op. cit.35. Stephen G. Walker, “The Motivational Foundations of Political Belief Systems: A Reanalysis of the

Operational Code Construct”, International Studies Quarterly, Vol. 27, No. 2 (1983), pp. 179–201; Walker,“The Evolution of Operational Code Analysis”, op. cit.36. Schafer and Walker, op. cit.37. Walker, “The Motivational Foundations”, op. cit.38. Walker, Schafer and Young, op. cit.

172 Leadership Analysis at a “Great Distance”

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

games to represent and predict different strategic situations by using Brams’Theory of Moves approach. Brams’ typology is generic enough to appeal toresearchers outside of North America.39 However, given the general unavailability

Table 1. Contents of the Revised Holsti Operational Code Typology.

TYPE AConflict is temporary, caused by humanmisunderstanding and miscommunication.A “conflict spiral”, based uponmisperception and impulsive responses, isthe major danger of war. Opponents areoften influenced in kind to conciliation andfirmness. Optimism is warranted, basedupon a leader’s ability and willingness toshape historical development. The future isrelatively predictable, and control over it ispossible. Establish goals within aframework that emphasises sharedinterests. Pursue broadly internationalgoals incrementally with flexiblestrategies that control risks by avoidingescalation and acting quickly whenconciliation opportunities arise.Emphasise resources that establish aclimate for negotiation and compromiseand avoid the early use of force.

Settle > Deadlock > Dominate > Submit

TYPE CConflict is temporary; it is possible torestructure the state system to reflect thelatent harmony of interests. The source ofconflict is the anarchical state system,which permits a variety of causes toproduce war. Opponents vary in nature,goals and responses to conciliation andfirmness. One should be pessimistic aboutgoals unless the state system is changed,because predictability and control overhistorical development are low underanarchy. Establish optimal goalsvigorously within a comprehensiveframework. Pursue shared goals, butcontrol risks by limiting means ratherthan ends. Act quicklywhen conciliationopportunities arise and delay escalatoryactions whenever possible. Resourcesother than military capabilities areuseful.

Settle > Dominate > Deadlock > Submit

Dominate > Settle > Deadlock > Submit

Conflict is permanent, caused by humannature(D), nationalism (E) or internationalanarchy(F). Power disequilibria are major dangersof war. Opponents may vary, and responsesto conciliation or firmness are uncertain.Optimism declines over the long run and inthe short run depends upon the quality ofleadership and a power equilibrium.Predictability is limited, as is control overhistorical development. Seek limited goalsflexibly with moderate means. Usemilitary force if the opponent andcircumstances require it, but only as a finalresource.

TYPE DEF

Dominate > Deadlock > Settle > Submit

Conflict is temporary, caused by warlikestates; miscalculation and appeasementare the major causes of war. Opponentsare rational and deterrable. Optimism iswarranted regarding realisation of goals.The political future is relativelypredictable, and control over historicaldevelopment is possible. One shouldseek optimal goals vigorously within acomprehensive framework. Controlrisks by limiting means rather thanends. Any tactic and resource may beappropriate, including the use of forcewhen it offers prospects for large gainswith limited risks.

TYPE B

Note: Instrumental beliefs are in bold, and philosophical beliefs are not.Source: Stephen G. Walker, “The Motivational Foundations of Political Belief Systems: A Reanalysis ofthe Operational Code Construct”, International Studies Quarterly, Vol. 27, No. 2 (1983), pp. 179–201;Stephen G. Walker, “The Evolution of Operational Code Analysis”, Political Psychology, Vol. 11, No. 2(1990), pp. 403–418.

39. Steven Brams, Theory of Moves (Cambridge: Cambridge University Press, 1994).

Özgür Özdamar 173

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

of training in quantitative political science and FPA outside of North America,game theory is rarely used in the academia of the MENA region.Contemporary operational code analysis uses an automated content analysis

method called the “Verbs in Context System” (VICS), introduced by Walker,Schafer and Young.40 This system focuses on the verbs in a leader’s public state-ments and their attributions regarding the exercise of power to self and others toconstruct quantitative indices that correspond to the philosophical and instrumen-tal beliefs shown in Table 2.41 The VICS indices for the master beliefs, P-1 (nature ofthe political universe), I-1 (strategic approach to goals) and P-4 (ability to controlhistorical development), are mapped on the vertical (P-1/I-1) and horizontal (P-4)axes in Table 1 to locate the leader’s images of self and other in one of the four quad-rants. The locations for self (I-1, P-4a) and other (P-1, P-4b) lead to predictionsregarding strategic preferences about the goals of settle, submit, dominate anddeadlock.42 Table 2 provides details about calculating VICS indices.Walker and Schafer further refined the preference orderings based on the key

operational code indices and developed a “theory of inferences about prefer-ences”.43 Inferences are made by comparing the leader’s key operational codescores (P-1, I-1 and P-4) with a norming sample of world leaders (see Table 3). Ifthe index scores lie above (below) this norm, the index is considered positive (nega-tive) for the purposes of inferring preferences. For the P-4 index, a norming range ofone standard deviation was set, and the index is interpreted by looking at whetherthe score is within one standard deviation of the norming mean or outside of it.44

In my research, due to a limited number of speeches in English, I chose toperform hand coding. Therefore, I could use Turkish texts and texts translatedinto Turkish from Arabic and Farsi. To simplify the research and thereforeachieve more reliable results, I coded only positive/negative attributes about self/other. These variables were enough to determine the operational code masterbeliefs: P-1, I-1 and P-4. As will be discussed later in this article, this simplerapproach achieved high levels of intercoder reliability.There is still an ongoing debate in the scholarly community about whether hand

coding has more advantages over the automated system.45 I suggest that there is atrade-off between validity and reliability of coding depending on the specificmethod used. Like many operational code analysis researchers who have con-ducted both computerised and hand coding, I argue that properly executed handcoding is significantly more valid than the computer-based analysis. The reasonis that even after two decades of development, Profiler Plus software is notcapable of coding verbs in context 100% correctly. This is not a problem specificto only Profiler Plus and operational code research. Other fields, such as in compu-ter science, also lack complete accuracy in natural language processing (NLP). After

40. Walker, Schafer and Young, op. cit.41. Schafer and Walker, op. cit.42. Ibid.43. Ibid.44. Ibid.45. Henk E. Goemans, Kristian Skrede Gleditsch and Giacomo Chiozza, “Introducing Archigos: A

Dataset of Political Leaders”, Journal of Peace Research, Vol. 46, No. 2 (2009), pp. 269–283; MargaretG. Hermann, Assessing Leadership Style: A Trait Analysis (Social Science Automation System, 1999);M. Schafer, “Issues in Assessing Psychological Characteristics at a Distance: An Introduction to the Sym-posium”, Political Psychology, Vol. 21, No. 3 (2000), pp. 511–527, available: <http://www.jstor.org/stable/3791848>.

174 Leadership Analysis at a “Great Distance”

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

Table 2. “Verbs in Context” Belief Indices in a Leader’s Operational Code.

P-1 NATURE OF THE POLITICALUNIVERSE (Image of others)

%Positive minus%Negative Transitive Other Attributions

+1.0 friendly to−1.0 hostile

P-2 REALISATION OF POLITICALVALUES (Optimism/Pessimism)

Mean Intensity of TransitiveOther Attributions divided by 3

+1.0 optimistic to−1.0 pessimistic

P-3 POLITICAL FUTURE(Predictability of others’ tactics)

1 minus Index of QualitativeVariation for Other Attributions

1.0 predictable to0.0 uncertain

P-4 HISTORICAL DEVELOPMENT(Locus of control)

Self (P-4a) or Other (P-4b) Attributionsdivided by [Self plus Other Attributions]

1.0 high to0.0 low self-control

P-5 ROLE OF CHANCE(Absence of control)

1 minus [Political Future×HistoricalDevelopment Index]

1.0 high role to0.0 low role

I-1 APPROACH TO GOALS(Direction of strategy)

%Positive minus%Negative Self Attributions

+1.0 high cooperation to−1.0 high conflict

I-2 PURSUIT OF GOALS(Intensity of tactics)

Mean Intensity of Transitive SelfAttributions divided by 3

+1.0 high cooperation to−1.0 high conflict

I-3 RISK ORIENTATION(Predictability of tactics)

1 minus Index of QualitativeVariation for Self Attributions

1.0 risk acceptant to0.0 risk averse

I-4 TIMING OF ACTION(Flexibility of tactics)

1 minus Absolute Value[%X Minus %Y Self Attributions]

1.0 high to0.0 low shift propensity

a. Coop v. Conf tactics Where X = Coop and Y = Confb. Word v. Deed tactics Where X =Word and Y =Deed

I-5 UTILITY OF MEANS(Exercise of power)

Percentages for exercise of powerCategories a through f

+1.0 very frequent to0.0 infrequent

a. Reward a’s frequency divided by totalb. Promise b’s frequency divided by totalc. Appeal/Support c’s frequency divided by totald. Oppose/Resist d’s frequency divided by totale. Threaten e’s frequency divided by totalf. Punish f’s frequency divided by total

Note: All indices vary between 0 and 1.0 except for P-1, P-2, I-1 and I-2, which vary between 1.0 and þ 1.0. P-2 and I-2 are divided by 3 to standardise the range.Source: Stephen G. Walker, Mark Schafer and Michael D. Young, “Systematic Procedures for Operational Code Analysis: Measuring and Modeling Jimmy Carter’sOperational Code”, International Studies Quarterly, Vol. 42, No. 1 (1998), pp. 175–190.

Özgür

Özdam

ar175

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

seven decades of research, computer scientists still struggle with creating efficientplatforms to help interactions between human languages and computers. Todaymost NLP tools are based on a set of handwritten rules (as is Profiler Plus).These tools are still not as accurate as humans when coding a specificallyfocused text such as in operational code studies analysis in political science.Perhaps this will change as computer-based analysis takes a more “machine learn-ing” direction. Especially in MENA political settings, where political speeches arevery colourful, witty, sarcastic, indirect and reproachful, researchers are better offresorting to hand coding to capture the details and meaning of speeches. Theadvantage of computer-based analysis is, of course, a maximum level of reliability.Because computers make coding based on the same set of rules each time, the sameprocedure is applied to each text regardless of culture, political bias or linguisticproblems. Computerised coding also allows for norming group comparisonbecause of the reliability. In this research, however, one cannot compare resultsto world leaders’ norming group scores due to reliability problems. Only resultsobtained from the automated VICS system are comparable to each other. Therefore,while I can compare these leaders to one another using my hand coded results, Icannot strictly compare them to the VICS norming group scores.

Case Selection

Political Islam is not a monolithic movement; neither is it static. Furthermore,countries’ historical backgrounds and political cultures have shaped thenature of the different movements. Yet, I argue that there are overarchingthemes and patterns of Islamist foreign policy beliefs and perceptions. Tocapture the diversity and the ideological similarities of the beliefs, I chose com-parable leaders in terms of the period in which they operated and with respectto the representation of different ethnicities. I present the reasons for thesechoices below.First, the three “old school” leaders (Erbakan, Qaddafi, and Khomeini) represent

the three major ethnicities in the Muslim MENA: Arab, Persian and Turkish.Second, they come from countries with very different political backgrounds.While Iran experienced the earliest example of constitutional monarchy in theregion, it went through the Pahlavi dynasty’s authoritarianism for about six

Table 3. Norming Scores and Inferences.

Self Other Values Preference Order in a 2×2 Strategic Game

I-1 & P-4a P-1 & P-4b + & > Settle > Deadlock > Submit > Dominate (Appease)I-1 & P-4a P-1 & P-4b + & = Settle > Deadlock > Dominate > Submit (Deter/Reward)I-1 & P-4a P-1 & P-4b + & > Settle > Dominate > Deadlock > Submit (Exploit)I-1 & P-4a P-1 & P-4b − & < Dominate > Settle > Submit > Deadlock (Bluff)I-1 & P-4a P-1 & P-4b − & = Dominate > Settle > Deadlock > Submit (Punish/Compel)I-1 & P-4a P-1 & P-4b − & > Dominate > Deadlock > Settle > Submit (Bully)

Note: “+” indicates above and “−” indicates below the norming mean; “<, >, =” indicate below, above orwithin the norming average range, which is P-4a ± 1 SD. The mean values for the norming group ofworld leaders are: P-l = +.30, SD = .29; I-1 = +.40, SD = .43; P-4 = .22, SD = .13.Source: Akan Malici and Allison L. Buckner, “Empathizing with Rogue Leaders: MahmoudAhmadinejad and Bashar al-Asad”, Journal of Peace Research, Vol. 45, No. 6 (2008), pp. 783–800.

176 Leadership Analysis at a “Great Distance”

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

decades, before the Islamic revolution overthrew the monarchy. Libya has neverexperienced democratic rule (except for the interim regime since the death ofQaddafi in 2011) and in modern history has been governed by the OttomanEmpire, Italy or local monarchs. Turkey, on the other hand, has enjoyed a multi-party democratic system since 1946, albeit interrupted by four military coupsbetween 1960 and 1997.

In terms of religious differences, Iran has been the only overtly Shia country inthe region, while Turkey and Libya are predominantly Sunni. Regarding the timeperiod, the three leaders were all active when Islamist leaders first came topower in the MENA, i.e. during the 1970s and 1980s.46 I only coded speechesfrom these two decades because the end of the Cold War changed the systemstructure as well as perceptions. Moreover, Khomeini died in 1989. The mostimportant reason for my selection of these three leaders is that they were overt Isla-mists and used their own mixtures of Islam and other schools of thought andnationalism.

In my selection of neo-Islamists, I use similar criteria. Erdoğan represents politi-cal Islam in a democratic country, namely Turkey. Ahmadinejad and Meshal alsoperform in some sort of electoral politics, yet one cannot really say that Iran andGaza are truly democratic. In terms of religious differences, Erdoğan and Meshalrepresent Sunni Islam and Ahmadinejad represents the Shia. The three leaders’ bio-graphies and their political/religious experiences are also strikingly similar.

Erdoğan was born in 1954 in Rize, a small city in north-eastern Turkey. Hebecame involved in Turkey’s rising political Islamist movement and metErbakan, by then a prominent political figure in the Islamic movement and whobecame the country’s first Islamist prime minister.47 In 1994, Erdoğan becamemayor of Turkey’s largest city, Istanbul. He established a new party (AKP) andbecame its leader in August 2001. Since then, Erdoğan has increased his share ofvotes, obtaining 49.8% of the popular vote in 2011 and 52% in 2014, in additionto becoming an important political figure abroad, especially in the MENAcountries.

Ahmadinejad was born to a middle class Persian family living in the village ofAradan in Semnan province in 1958.48 He studied civil engineering at the Iran Uni-versity of Science and Technology in Tehran. His early political life began withstudent movements in Tehran. In the 2005 general elections, he was the strongestconservative candidate for the presidency of Iran, garnering the support of theSupreme Leader Ali Khamenei. Following a populist election campaign, he took62% of the votes and became the sixth president of the Islamic Republic of Iran.During his first term in office, he followed an aggressive foreign policy againstIsrael. After a controversial general election in 2009, he once again was electedbut he failed to secure presidential nomination by the clergy in the 2014 elections.Ahmadinejad is considered to be one of the new conservatives in Iranian politicsand the MENA, therefore fitting the aim of this analysis to analyse later represen-tatives of political Islam in the region.

46. Qaddafi came to power in 1969 with a military coup. Erbakan was appointed deputy prime min-ister in 1973. Khomeini came to power with the Islamic revolution in 1979.47. “Profile: Recep Tayyip Erdoğan”, BBC (18 July 2007), available: <http://news.bbc.co.uk/2/hi/

europe/6900616.stm> (accessed 22 March 2012).48. “Iran’s President Launches Weblog”, BBC (14 August 2006), available: <http://news.bbc.co.uk/2/hi/

middle_east/4790005.stm> (accessed 18 March 2012).

Özgür Özdamar 177

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

Finally, Meshal was born in 1956 in Ramallah, Palestine. In 1967, after the Israelioccupation, his family moved to Kuwait, a hotbed of pro-Palestinian activism andArab nationalism during the 1970s.49 Meshal became involved in the Islamic andPalestinian causes as well as in Arab nationalism at an early age, and became amember of the Muslim Brotherhood in 1971. Later, while he was studyingphysics at Kuwait University, he established a radical student group known asthe “List of the Islamic Right”, in which he orchestrated a robust campaignagainst Yasser Arafat’s Palestine Liberation Organization (PLO).50 Meshal wasstill leader of this group when the Palestinian resistance movement Hamas wasfounded in 1987 after the Intifada, the first Palestinian uprising against Israeli occu-pation. Hamas aims to end the Israeli occupation of Palestinian territories and unifyformerly invaded territories under the authority of one Palestinian state in confor-mity with the 1967 borders. Meshal became chairman of the Hamas PoliticalBureau in 1996, which criticised Arafat for his weakness and inertia vis-à-visIsrael’s aggression and oppression of the Palestinian community. In 1997, Meshalwas the target of an assassination attempt authorised by Israel’s Netanyahuadministration, which described Meshal as the leading figure in Hamas andblamed him for the murder of innocent Israeli civilians in the Palestinian suicidebombings.51 In 1999, King Abdullah closed all the Hamas offices in Jordan andMeshal was imprisoned, and deported shortly thereafter.52 He is still the politicalchief of Hamas.These brief introductions to the three neo-Islamist leaders’ lives show significant

similarities. All three are from modest backgrounds, college educated, religiouslyraised, and have been active in and around religious institutions. All three roseup against the established order in their respective countries and used religion asa significant aspect of their political mobilisation tactics.

Research Design

I used a collection of speeches, book chapters, books written by the respectiveleaders, and press conferences from the late 1960s to 1990 (but mostly from the1970s and 1980s) to measure the operational codes of Erbakan, Khomeini andQaddafi. I coded 14 different sources. Erbakan’s speeches are taken from a compi-lation of his important speeches on foreign policy issues. Khomeini’s speeches aretaken from the Islamic Republic of Iran’s website on Khomeini and a book of hisspeeches from the early 1960s until 1981. Finding Qaddafi’s original speechesfrom this period proved difficult. I used a book that combined a rather long theor-etical treatise on capitalism and colonialism, Qaddafi’s “third universal theory” andhis (long) interviews with international journalists and academics from 1973. Allmaterial concerns foreign policy issues, international relations or history ingeneral, or the political economy of Muslim countries, focusing on foreign

49. “Profile: KhaledMeshaal”, Al Jazeera (18 April 2008), available: <http://www.aljazeera.com/news/middleeast/2008/04/20086150597219693.html> (accessed 22 March 2012).50. “Profile: Khaled Meshaal of Hamas”, BBC (8 February 2006), available: <http://news.bbc.co.uk/2/

hi/middle_east/3563635.stm> (accessed 22 March 2012).51. “Profile of Khaled Meshaal”, Council on Foreign Relations (13 July 2006), available: <http://www.

cfr.org/terrorist-leaders/profile-khaled-meshal-aka-khalid-meshaal-khaleed-mashal/p11111> (accessed22 March 2012).52. “Profile: Khaled Meshaal of Hamas”, op. cit.

178 Leadership Analysis at a “Great Distance”

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

economic relations. To measure Erbakan’s operational code, I coded six relativelylong speeches and detected 413 self/other attributions. For Khomeini, I found 352attributions in five speeches. For Qaddafi, in two long chapters that includedmore than 20,000 words, I found 284 verbs and other attributions. All speeches con-tained at least 20 coded verbs, as suggested by Schafer and Walker.53

Ahmadinejad’s speeches are from the second half of the 2000s, with four fromUN General Assembly meetings and three from interviews with journalists. Icoded 286 self/other attributions from a total of 17,141 words. The sources forMeshal’s speeches are mostly interviews with international media, which includeda total of 13,929 words. There are seven Meshal speeches coded, with a total of 349attributions. Erdoğan’s speeches are taken from the AKP website. In a total of 10speeches (comprising 21,932 words), 372 negative/positive and self/other attribu-tions were coded. Because I coded all sources by hand, the results cannot be com-pared to the computer-based norming group scores,54 though they may becompared to one another in the hand coded set. This issue will be discussed inthe next section.

The title, source and language of each speech, and the results obtained fromthem, are provided in an appendix. The speeches were selected according to criteriaproposed by Walker et al. (each speech is at least 1,000 words and includes at least20 transitive verbs) and their availability.55 Two large libraries in Ankara (BilkentUniversity Library and the National Library of Turkey), relevant embassy librariesin Ankara (such as the Iranian and Libyan embassies) and internet sources weresearched and all the available speeches that fulfilled Walker et al.’s criteria56 havebeen used. Only in the case of Qaddafi did the number of speeches not fit the cri-teria due to a lack of his available speeches. However, I have included Qaddafibecause the volume of two book chapters written by him exceeds the word limitcriteria suggested by Walker et al.57 All speeches were coded by the author andtwo research assistants, and intercoder reliability checks were conducted.

Results

The analysis reveals interesting and consistent results about Islamist leaders’ oper-ational codes. Tables 4 and 5 present numeric results from hand coding of the sixleaders’ speeches. The three original Islamist leaders’ self-images fall under quad-rant A, as Figure 1 shows. That is, Erbakan’s, Qaddafi’s and Khomeini’s self-images range from definitely cooperative to very cooperative, with low historicalcontrol.58 All three leaders’ self-images are clearly within the Type A category,with a high cooperative outlook and personal sense of control. Type Aleaders have a relatively friendly strategic orientation and a relatively low sense

53. Schafer and Walker, op. cit.54. Ibid.55. Walker, Schafer and Young, op. cit.56. Ibid.57. Ibid.58. In many operational code studies, a norming group comparison has not been made until very

recently. Instead researchers relied on categories based on absolute operational code scores. Forexample, P-4 categories are very low (at 0), low (at 0.25), medium (at 0.50), high (at 0.75) and veryhigh (at 1.0). Every score in between is reported as an interval. For example, if a leader’s P-4a score equal-led 0.40, this would be reported as “low to medium” (Walker, Schafer and Young, op. cit.).

Özgür Özdamar 179

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

of historical control. Such leaders tend to exhibit choice and shift propensities thatfavour tactics of “appease” and “bluff”, associated with an appeasement strategy59

(see also Table 1 for further explanation).The operational code analyses of the three former Islamist leaders exhibit some

further uniformity. Erbakan’s, Qaddafi’s and Khomeini’s perception of the“other” is in quadrant B. That is, other leaders and the political world aresources of conflict, and historical control lies with the opponent. Type B leaders

Table 5. Master Beliefs Scores for “Neo-Islamists”: Erdoğan, Ahmadinejad and Meshal.

P-1 I-1 P-4a P-4b

Erdoğan −0.73 0.90 0.46 0.54Ahmadinejad −0.83 0.84 0.23 0.77Meshal −0.95 0.80 0.28 0.72

Figure 1. Three Original Islamist Leaders According to Their Leadership Type.

Table 4. Master Beliefs Scores for “Original Islamists”: Erbakan, Qaddafi and Khomeini.

P-1 I-1 P-4a P-4b

Erbakan −0.82 0.8 0.34 0.66Khomeini −0.89 0.44 0.24 0.76Qaddafi −0.95 0.62 0.28 0.72

59. Stephen G. Walker, “Forecasting the Political Behavior of Leaders with the Verbs in ContextSystem of Operational Code Analysis”, Research Report (2000).

180 Leadership Analysis at a “Great Distance”

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

exhibit a relatively hostile strategic orientation and a high sense of historicalcontrol. They are likely to favour tactics of “bully” and “exploit”, associated witha brinkmanship strategy.60 Considering the background of Islamism as a reactionto colonialism and Western penetration into the region, these results show the val-idity of the coding because they represent high levels of distrust and “fear” ofothers.

The neo-Islamist leaders’ operational codes also feature similarities, as Table 5and Figure 2 show. My analysis of Erdoğan’s speeches reveals results that areboth consistent with and challenges for conventional images of Islamists. Accord-ing to the analysis, Erdoğan believes that the political universe is very hostile andthat he has medium control over historical developments. According to Erdoğan’soperational code, others in the political universe pursue a relatively hostile strategicorientation and have more historical control over events (Type B). They likelyfavour a brinkmanship strategy.61 Erdoğan’s conceptualisation of himself is inquadrant A. He pursues shared interests by following flexible strategies andshuns brinkmanship as much as possible. According to Walker, Type A leaderspursue a relatively friendly strategic orientation but feel that they have relativelyless historical control over events. These leaders presumably exhibit choice andshift propensities.62

Ahmadinejad’s speeches also reflect a conflictual worldview. He perceives his-torical control as belonging with the other. This propensity appears to be atypical feature of Iranian politicians, given their historical problems with foreignpowers. Ahmadinejad also sees himself as very cooperative. Meshal appears tohave the most negative worldview. His P-1 (sense of the political universe as con-flictual or cooperative) score is almost as conflictual as it can get. Given the status of

Figure 2. The Neo-Islamist Leaders According to Their Leadership Type.

60. Ibid.61. Ibid.62. Ibid.

Özgür Özdamar 181

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

the Palestinian Authority and its relations with Israel and the rest of the world, thisperception is understandable. Meshal also seems to be pro-cooperation regardinghis in-group, and his low level of historical control is in line with political Islamists’worldviews and belief systems. Both Ahmadinejad’s and Meshal’s self-perceptionsare Type A while they perceive the other as Type B.My analysis of the philosophical and instrumental beliefs of the three original

Islamist leaders reveals some patterns in how political Islam as an ideologyappeared and articulated itself. The most general theme in these leaders’ speechesis an anti-West sentiment. For Qaddafi, this was reflected most frequently as anti-colonialism because of Libya’s experience as a colony; for Khomeini, it manifestedas anti-imperialism because of how Iran was pushed and pulled by the greatpowers of the 19th and 20th centuries; and for Erbakan, it was revealed as anti-international capitalism, perhaps due to Turkey’s economic decline during the18th and 19th centuries and its reduction to small power status by Britain andFrance after World War I. The original three Islamist leaders were very critical ofall forms of Western penetration into the region. Their criticism and high levelsof distrust of the West63 are confirmed by the operational code approach. Allthree leaders exhibited very negative and high P-1 scores.Such a historical background also contributes to these leaders’ uniform percep-

tions of historical control lying more with the “other” than the self, as P-4a(sense of mastery or control) scores suggest. Colonisation, imperialism and greatpower penetration into their countries and the region, as well as the gradual weak-ening and final destruction of the Ottomans by Western powers, diminishedregional leaders’ self-confidence in general. Specifically, due to their ideology andhow they analyse the region’s history, Islamist leaders are not fully confidentabout historical control compared with Western leaders whom operational coderesearchers have analysed.64 In all speeches and articles coded, there is an extre-mely high level of historical control attributed to others. Political Islamistsmostly do not attribute social, political and economic changes to domesticfactors. The high scores of P-4b and low scores of P-4a,65 mentioned above, arethe patterns detected in political Islamists’ operational codes in addition tosimilar P-1 (political universe as cooperative or conflictual) and I-1 scores (whichrepresents instrumental beliefs of a leader in terms of cooperation and conflict).Finally, in terms of the strategies employed, the “original” political Islamists

present a mixed picture. In the most general pattern, they see themselves as coop-erative if opportunities arise. Towards their in-group, all six leaders seem coopera-tive. However, how the in-group is defined varies from one Islamist to another.While some Islamist leaders (such as Erbakan, Erdoğan and Ahmadinejad) per-ceive all Muslims as their in-group, the in-groups of some other leaders havebeen very small, such as Khomeini’s in-group which included only Shia clergyand oppressed Palestinian Muslims. The new Islamists show similar categoricalbeliefs to each other in terms of the P-1, P-4 and I-1 constructs.

63. Black, op. cit.64. For example, Jimmy Carter’s P-4 score was 0.65 during 1977 to 1979 (Walker, Schafer and Young,

op. cit.); or Margaret Thatcher’s varied in her tenure depending on the issue between 0.58 and 0.73(Crichlow, op. cit.). The Islamist leaders’ P-4 varied from 0.23 to 0.46.65. P-4 (mastery or control of history) is calculated as P-4b = 1–P-4a. Therefore, low scores of P-4a

(self’s mastery) mean higher scores of P-4b (other’s mastery).

182 Leadership Analysis at a “Great Distance”

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

This consistency in scores over time periods and between leaders seems toconfirm that political Islam as a distinct political ideology seems to have a ratherunified worldview about international affairs. On the other hand, my expectationsof more significant differences existing between former and current Islamist leaderswas not supported in this preliminary analysis. Only Erdoğan’s historical control(P-4a) is significantly higher than that of the other Islamist leaders; he is also gen-erally friendlier towards others, including the West. This analysis shows generalpatterns of beliefs about foreign policy among Islamist leaders. The operationalcodes of the second three representative Islamist leaders show similar approachesto foreign policy, despite their different personal and country-specific backgrounds.It appears that operational code analysis is a theoretically robust construct that canbe applied to other political units, regions and cultures. However, methodologi-cally, researchers must be very careful about the effects of culture and geographyon results. The boundaries of operational code research methodology may not beextended to other cultures and regions easily, which is the topic of the next section.

Theoretical and Methodological Issues Concerning the Use of OperationalCode Analysis to Analyse Middle Eastern and North African Islamist Leaders

Theoretical Issues

Using the operational code construct to analyse Islamist leaders in the MENA hastheoretical justifications. To begin with, the construct is one of the most theoreti-cally developed leadership analysis tools in the literature. It has been applied tomore than 30 world leaders, with many essays on the subject published inrespected scientific journals. Therefore, it has a proven record of prediction andexplanation. The researcher can resort to many different studies and learn fromother researchers’ experiences. Furthermore, operational code analysis was devel-oped to analyse a group of political decision-makers’ belief systems with the aimof determining patterns and making behavioural predictions; this is exactly theaim of my research on political Islamists. In short, the level of theoretical develop-ment of operational code analyses is one of the major reasons I chose the oper-ational code construct to analyse MENA leaders.

Using the operational code construct in a new context has been discussed pre-viously in the literature. For example, Alexander George66 simplified and theoreti-cally codified the operational code construct for the aim of applying it to otherpolities than the USSR.

I have tried in this paper to codify the general issues and questions aroundwhich such a belief system is structured in the hope that it will encourageand facilitate systematic efforts to apply this research approach to a varietyof other ruling groups and individual political leaders as well. The possi-bility emerges of a useful new dimension for comparative studies of differ-ent leaders and elite groups.67

66. Alexander L. George, The “Operational Code”: A Neglected Approach to the Study of Political Leadersand Decision-Making (Santa Monica, CA: RAND Corporation, 1967), available: <http://www.rand.org/pubs/research_memoranda/RM5427.html>, also available in print form: George, “The OperationalCode”, op. cit.67. Ibid., p. 220.

Özgür Özdamar 183

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

George also takes up this challenge in the article by introducing brief analyses ofcases from American foreign policy while discussing the operational codeconstruct.StephenWalker also discussed the applicability of the operational code construct

in a comparative manner.68 Walker, when discussing the flourishing of the oper-ational code literature after George’s seminal article in 1969, suggests that the fol-lowing two decades witnessed many US scholars using the construct to analysevarious American decision-makers. Walker discusses the difficulties associatedwith converting the operational code construct from the Soviet focus to USdecision-makers.69 He suggests that only some research has successfully madethe theoretically consistent link between US leaders’operational code and their per-ception of the Soviet Union by using both quantitative operational code analysisand process tracing techniques. In a later article in which Walker applies Lakato-sian criteria of scientific progress to the research programme, he proposes a cau-tionary optimism about the success of operational code analysis in terms ofscientific progress.70 Although overwhelming evidence suggests that “beliefsmatter” in general, the operational code construct can only be said to show scien-tific progress if one can prove that beliefs do not fall into the endogeneity trap, i.e.they are independent from the structure’s effects. For example, when tested againstevidence (by analysing many different leaders from many different countries), theoperational code construct can account for why European leaders gave completelydifferent responses to Germany before 1914 and 1939 despite structural variablesbeing quite similar. The answer is the differences in personal beliefs and theagent’s interpretation of the structure, which the operational code construct canaccount for.Operational code analysis also fits well with the nature of the data on MENA

leaders. In many countries, the quality of bureaucracy is rather low, with, forexample, few services for recording, archiving and publishing leaders’ speeches.This may be because such countries come from a so-called “oral tradition”,where state records, speeches and even literary works are less often and less pro-fessionally recorded. Furthermore, MENA countries have scant academic literatureand no journalistic tradition of psycho-biographies. For these reasons, using aleader’s available speeches for analysis at a distance is the best choice.The main problem I encountered during my research was a classic “level of

analysis” problem; in this case, single-leader analysis versus an (elite) group-of-decision-makers analysis. Operational code analysis can be applied to leaders aswell as groups. In some cases, leader-focused analysis is illuminating in terms ofexamining a country’s foreign policy behaviour. If a country (such as the UnitedStates) has a presidential system, or a strong leader is in power, leader-focusedanalysis is perhaps more appropriate. In other polities, analysing a certaingroup’s impact on foreign policy decisions is more important. In such cases, theadministrative and political systems of the country assign a greater role togroups rather than single leaders. Therefore, while conducting the analysis I

68. Walker, “The Evolution of Operational Code Analysis”, op. cit.; Walker, Forecasting the Political Be-havior, op. cit.69. Walker, “The Evolution of Operational Code Analysis”, op. cit., p. 414.70. Stephen G. Walker, “Operational Code Analysis as a Scientific Research Program: A Cautionary

Tale”, in Colin Elman and Miriam Fendius Elman (eds.), Progress in International Relations Theory(London: MIT Press, 1998), pp. 245–276.

184 Leadership Analysis at a “Great Distance”

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

realised that the predictions one can make from operational code analysis must bereviewed cautiously. If a country’s foreign policy decision-making system is verycomplicated (as is Iran’s), predicting that country’s foreign policy decisions onthe basis of operational code analysis should perhaps be undertaken more cau-tiously. For example, group analysis would be more relevant for Hamas becauseits structure is not as homogenous as an organised Western political unit, such asa cabinet. The foreign policy behaviour of the group is shaped by different units,such as sub-groups or individuals. That said, I did not have a major problemwith the unit of analysis because my aim was to analyse different leaders from acertain political ideology and make inferences about that ideology. However, ifone wanted to predict Iranian foreign policy behaviour, one would have to bemore careful not to over-generalise the results.

Furthermore, during the research, I realised that mere operational code analysismay not prove sufficient to predict a country’s foreign policy behaviour at everydecision point. Depending on the system, constitutional rules and the decisionitself, different decision units may be more important.71 In Turkey, for example,in most cases the foreign policy decision unit is the cabinet, which usuallyfollows the prime minister’s choices. However, the Turkish constitution requiresa parliamentary decision to go to war and to admit foreign troops into thecountry. This difference explains why, although Erdoğan was in favour of joiningthe Iraq War in 2003, the Turkish parliament rejected that policy and did notadmit US troops into Turkey.72

Another theoretical issue about the boundedness of operational code analysisregards contextual or temporal limitations. In many cases in the MENA, it islikely that a state’s foreign policy agenda is dominated by a specific problem.Thus, the leader’s foreign policy belief system is shaped by the set of specific andparamount events of that period. In such cases, it becomes difficult to determinethe leader’s general foreign policy orientation.73 For example, the KurdistanWorkers’ Party (PKK) issue continues to dominate the Turkish foreign policyagenda and therefore the content of Turkish leaders’ foreign policy speeches.74 Insuch cases, the narrow focus on foreign policy speeches leads to distorted oper-ational code results in terms of explaining the perception of the leaders andgeneral foreign policy tendencies of these countries. Likewise, I faced similar pro-blems when analysing Hamas leaders’ operational codes. The conflict with Israeland the Israeli occupation of Palestinian territories constitute the most vital itemsin Hamas’s foreign policy agenda. Therefore, the salience of the conflict betweenthe Israeli and Palestinian parties makes it extremely difficult to differentiatebetween the operational code of the Hamas leadership on general foreign affairsand between Israel and the status of Palestine. This requires finding differentsources of speeches and articles and examining them carefully.

71. Margaret G. Hermann, “How Decision Units Shape Foreign Policy: A Theoretical Framework”,International Studies Review, Vol. 3, No. 2 (2001), pp. 47–81.72. Zeynep Taydaş and Özgür Özdamar, “A Divided Government, an Ideological Parliament, and an

Insecure Leader: Turkey’s Indecision about Joining the Iraq War”, Social Science Quarterly, Vol. 94, No. 1(2013), pp. 217–241.73. There is a debate about whether leaders have a general operational code or whether there are sep-

arate operational codes for different issues of foreign policy (Walker, Schafer and Young, op. cit.).74. The PKK is an armed militia group that has fought against Turkish governments since 1984 to

secede south-east Turkey in order to establish an independent Kurdish state. The PKK is on the list ofterrorist organisations for Turkey, the EU and the United States.

Özgür Özdamar 185

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

Operational code analysis can result in several insights and the ability to elicitgeneral predictions about the decision-making unit under discussion. Yet theseresults remain too abstract when not evaluated within a broader researchagenda. Operational code findings must be incorporated into different literaturessuch as game theory and/or leadership trait analysis. The preference order of aleader found by operational code analysis can be an effective starting point forestablishing a game in which the leader’s preferences can be plugged into agame theoretical model that focuses on a particular foreign policy decision.75 Theleadership typology literature is also an appropriate sphere in which to evaluatethe operational code of the leader in terms of locating him or her within a specifictypological model.76

However, the problem with these cross-theoretical endeavours is that theyrequire additional training. Considering the modest methodology course require-ments in PhD programmes throughout the world, except for the United States, itbecomes difficult to co-author with other scholars or even one’s own graduate stu-dents. Therefore, the theoretical richness and eclecticism of operational coderesearch becomes an obstacle even to itself. Many researchers are cautious aboutlearning to undertake this analysis because they feel they cannot publish theirresults without using additional tools and literatures. Learning operational codeanalysis requires a significant investment of time and energy, and theoretical andmethodological eclecticism works against drafting more analysts into thediscipline.

Procedural and Methodological Problems

Perhaps the most pressing issue with the operational code construct is related tolanguages. When coding Turkish texts and texts translated into Turkish fromArabic and Farsi, it was obvious that the logic of operational code analysis isbased exclusively on English. Sentences in English texts are usually relativelyshort, sentence structures are straightforward, and use of the passive voice isminimal. This kind of simple yet efficient writing seems exclusive to English;French and German political and academic literatures seem unnecessarily compli-cated. The situation is similar for Turkish, Farsi and Arabic.Middle Eastern languages are ancient and rich in their expressions. Because

speeches are generally designed for leaders to outsmart their political rivals byappearing cleverer and wittier, they are very difficult to decompose and codeaccording to operational code rules. For example, in political speeches intendedto excite public rallies, there are countless sentences without a verb. These sen-tences are nonetheless very strong politically and exhibit negative attitudestowards foreign powers or domestic political rivals. At this point, the researcherfaces a dilemma: whether to code this obviously negative attribute against thepolitical other or abide by the operational code protocol and exclude any sen-tence without a verb (or transitive verb). In my research, I tried to do theformer. I relaxed the coding rules in favour of non-transitive verbs and designeda new and consistent protocol to code Turkish and texts translated from Arabic

75. Walker, Forecasting the Political Behavior, op. cit.76. JerroldM. Post (ed.), The Psychological Assessment of Political Leaders. With Profiles of Saddam Hussein

and Bill Clinton (Ann Arbor, MI: University of Michigan Press, 2005).

186 Leadership Analysis at a “Great Distance”

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

and Farsi into Turkish. In the end, our coding—i.e. that of the principalresearcher and two research assistants—showed very high levels of intercoderreliability.

Coding transitive verbs exhibits a similar problem.77 In English and non-Englishspeeches, transitive verbs are coded as reflections of negative or positive attribu-tions to self and other. Yet it does not suffice to code negative and positive attribu-tions associated with a verb only; sometimes the context of the sentence implies aclear negative attribution even though the sentence does not include a transitiveverb (as in the verb-less sentences discussed above). For example, NecmettinErbakan said: “They [Americans] are the greatest pharaohs to the Muslimpeople”. According to the original operational code protocol, this sentence hasno coding value. However, in hand coding, if one understands the political ideol-ogy of the owner of this sentence, the culture of the country, the biblical reference topharaohs and its meaning for political Islamists today, one would know this sen-tence means: “Americans oppressed (or persecuted) the Muslim people morethan anybody else did”.

Decomposing and rewriting the sentence as such, the researcher can code it as“Other-Negative” and not miss such a clear and strong negative attribution toother. Automated systems are not capable of doing this. There are also othertypes of verbs in other languages that can refer to deeds and actions but that donot exist in English (for example, işteşfiiller—widely used in Turkish—is roughlythe same as reciprocal verbs in English, where the verb refers equally to two sub-jects, e.g., as in “exchange” and “marry”). In Turkish, these verbs may also refer tostrong political actions (negative or positive) that should be included in the coding.For these reasons, I believe that hand coding may be a better option for analysingspeeches in different political contexts and languages. I also maintain that richexamples of adjectives or adjectives used as verbs should also be examined for attri-butions of self and other. Furthermore, the researcher must decide how or whetherto code rich sets of expressions, proverbs, aphorisms, idioms and adjectives thatpoliticians speaking in MENA languages frequently use.

For at least two reasons, hand coding is also better at differentiating between apolitician’s in-group and out-group. First, in automated systems, the computercodes according to “I” and “we”. In the political culture of the Middle East, “I”and “we” are often reversed. Like the “royal we” in English, politicians like touse “we” when they mean “I”. Culturally, the idea is to appear more humble bynot using the “I”, but it leads to some confusion for the researcher. Second,leaders’ in-groups and out-groups can be completely different even if they hailfrom the same ideology; automated systems’ focus on pronouns may not givethe same results as hand coding. For example, for Erbakan all Muslims were hisin-group and therefore his approach was very cooperative. Khomeini’s in-groupwas much smaller, however; to him, the only groups that deserved his cooperationwere the Shia clerics and some oppressed Muslim communities, such as Palesti-nians. I believe that hand coding is the best tool to capture these differences and

77. Another smaller problem is the minimum number of verbs necessary to code. The general prin-ciple in operational code analysis is to code all speeches that include at least 20 transitive verbs.Having such a threshold was sometimes not practical. Although the number of verbs was low insome speeches, they nonetheless included strong and determinant attributions that reflected itsgeneral theme. Perhaps it is best to let the hand coder determine the eligibility of such speeches on anindividual basis.

Özgür Özdamar 187

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

thus improve the coding’s validity. If, for example, a leader’s mind is not as trans-parent as an American president’s, the traditional operational code rules do notapply. If one is coding a speech from Richard Nixon or Lyndon B. Johnsonduring the height of the ColdWar, “I”, “we” and “us” are rather easy to determine:meaning themselves, the West, NATO and/or American allies. But for Khomeini,the West was not the only enemy; Muslims who envied the West, Muslimswho allied with the United States, and Sunnis in general also belonged to theout-group.

Problems Related to MENA Politics

The most specific problem in applying operational code analysis to MENA leadersis that political Islam as a movement is diverse and non-bounded. When I beganthis research, I had the following problem in mind: Leites had used operationalcode to analyse a narrow group of people. The smaller group size (elite Bolsheviks),its unified ideology (Leninism), specific institutional structures (decisions made bythe Politburo) and the members’ cultural and linguistic similarities (all Soviet citi-zens speaking Russian) rendered his analysis capable of making rather soundgeneralisations.78

For political Islamists, except for theQuran and the Sunna, the ideological sourcesthey read are much more diverse than the Politburo members’ sources were. Thesources of Islamist leaders’ issues and the political influences on them are alsomore diverse. For example, Khomeini’s political views were strictly based on Shiareadings and theological ideas about the Twelve Imam traditions of Shiism.Erbakan, on the other hand, seemed to have been affected by Sunna insights(deeds of the Prophet Mohammed) and some Turkish and Kurdish religious scho-lars such as SaidNursi or those of theNaksebendi order. Given the region’s diversityof religious sects and beliefs, countries’ historical experiences and the various reli-gious and philosophical references, I questioned whether I could generalise abouta set of leaders. However, when the results showed great consistency, operationalcode analysis seemed a good choice for analysing even such a diverse group.As noted earlier, data availability was a major problem. Specifically, speeches by

Qaddafi were difficult to find due to political reasons (authoritarianism in thecountry and Libyan foreign service employees’ fear of possible consequences ifthey gave me access to speeches). Due to its loose institutional structure and itsabove-discussed focus on one foreign affairs issue, it was also difficult to accessdata that reflected Meshal’s foreign policy beliefs. In more developed systems, itwas much easier to access sources; one can easily download Erdoğan’s manyspeeches from the AKP’s website.

78. Using the operational code construct in a similar fashion to Leites’ handling of the Bolshevik partyraises a question about whether Bolsheviks and Islamists share similar approaches to politics. Theanswer is mixed. Bolsheviks’ ideology is associated with economic planning, communist state andsociety, atheism, one-party state and proletariat dictatorship and vanguardism, which are incompatiblewith Islamist ideology. By definition, Islamists are not atheists and they do not presume a utopian ega-litarian society. However, there are some interesting similarities between the two ideologies. Both ideol-ogies claim to be anti-imperialistic, and their discourse is generally anti-West for different reasons. Theyare both utopian in terms of the state and society they aspire to reach; in Islam this is a state representedby Prophet Mohammed’s brief tenure as a ruler of the first Muslim tribes in the 7th century whereas Bol-shevik ideology envisioned a stateless communistic society. Finally, both ideologies are cosmopolitan innature and claim to have transnational appeal that goes beyond borders.

188 Leadership Analysis at a “Great Distance”

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

Moreover, a general problem in operational code research with specific referenceto MENA is whether the operational code approach’s assumptions that courses ofactions and leader accountability are adequately reflected in leaders’ operationalcodes are correct. Operational code research assumes that leaders’ speechesreflect their belief systems; thus, one can predict their foreign policy behaviour.The operational code analysis literature also assumes a principle of accountability,which means that leaders’ actions are limited by what they say in their publicspeeches. This situation may be more the case in developed democracies, such asthe United States, where political accountability is expected from leaders. In suchcountries, leaders are held accountable for consistency between their speechesand actual foreign policy decisions. On the other hand, in the MENA, the principleof accountability is often lacking. Leaders may exhibit different foreign policy be-haviour than suggested from their public speeches. In autocratic regimes especially,leaders are much less accountable in all policy areas, including foreign policy.

There is also inconsistency between speeches and acts in domestic and foreignspeeches. For example, I realised that the tone of leaders’ speeches changes fromaudience to audience and subject to subject. Qaddafi seemed less aggressivetalking to French journalists than to the Libyan youth at home. When acting oninternational platforms, such as the United Nations, the leaders’ speeches seemmore “balanced” than their campaign speeches. For these reasons, the operationalcode rule that speeches must be about the same subject and addressed to similaraudiences must be implemented. We should be collecting as many speeches aspossible to prevent data distortion when analysing domestic and internationalspeeches.

The last regional difficulty involves the MENA’s lack of developed universitysystems, scientific production and theoretical foreign policy analysis research.There are few scholars—outside of North America—who focus on theoreticalapproaches coming from the periphery, the non-Western international relations(IR) community’s sociology, or scientific, quantitative studies of middle orsmaller powers’ foreign policies.79 The dearth of such resources meant that therewas little or no literature to resort to when encountering problems. Scientific experi-ence means learning from other researchers’ experiences, insights and mistakes.Yet, since there are not many FPA studies from the regional universities or focusingon the MENA, I lacked resources to consult when facing a problem.

Conclusion

A general evaluation of the above discussion suggests that the operational coderesearch agenda has great scientific value and is capable of producing insights

79. See Ali Karaosmanoğlu and Ersel Aydınlı, “Launching All Azimuth”, All Azimuth, Vol. 1, No. 1(2012), pp. 5–7 for an elaboration of the Western orientation of IR theories and the need to give voiceto alternatives from the periphery; Seçkin Köstem, “International Relations Theories and Turkish Inter-national Relations: Observations Based on a Book”, All Azimuth, Vol. 4, No. 1 (2015), pp. 59–67 for theTurkish IR community’s sociological analysis; Hong-Cheol Kim and Patrick James, “The Paradox ofPower Asymmetry: When and Why Do Weaker States Challenge US Hegemony?”, All Azimuth, Vol.5, No. 2 (2016), pp. 5–28 for a quantitative analysis of small state foreign policy; and Musa Tüzünerand Gonca Biltekin, “A Pilot Study of Quantifying Turkey’s Foreign Affairs: Data Generation, Challengesand Preliminary Analysis”, All Azimuth, Vol. 2, No. 2-SI (2013), pp. 45–70 for an analysis of Turkishforeign policy as a middle power.

Özgür Özdamar 189

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

when applied even to regions and political cultures that are very different from theUnited States. I believe the operational code construct is universally applicablebecause its foundational principles are based on universal psychological assump-tions. The success of operational code research reflects the success of cognitivelyoriented research paradigms in general. The operational code construct, as a viamedia position in the FPA field, i.e. the combination of competing cognitive researchand game theoretic models, is also appreciated.80 By including both competingparadigms in FPA, the operational code construct has gained insight and rigour.However, applying the operational code’s procedural and methodological

assumptions to non-Western groups or leaders must be carefully reconsidered.The experience from the MENA shows mixed results. Although I argue that thebasic logic of operational code research is valid for Middle Eastern and NorthAfrican political systems and leaders (and perhaps for all regions), details abouthow to produce operational code scores for leaders may need to be tailored accord-ing to the language used, characteristics of the political system, and the country.Above all, the coding rules designed for the English language may not provevalid when applied to other languages. The fine details, nuances and idiosyncrasiesof a certain language must be considered, and coding flexibility to overcome thesedifficulties should be tolerated by the operational code research community ingeneral. In some regions and cultures, such as the MENA, I suggest employinghand coding rather than running translated English speeches through automatedcoding.Adapting North American FPA theories to other cultural settings will benefit the

theory itself, the FPA field in general, and research and scientific developmentabout the country in question, as well as produce cumulative knowledge. As it isapplied to different leaders from different polities, operational code research willbe enriched. It will expand and become methodologically more diverse and eclec-tic, thus producing a more comprehensive body of knowledge. Such research willconsolidate FPA’s position as a subfield of IR, or as “the ground of IR”, as ValerieHudson suggests.81 FPA has the potential to serve as a bridge between theoryand practice, the abstract and the applicable, science and policy, the material andthe ideational.82 Theoretically and methodologically, it holds a bridging positionbetween the most abstract IR studies and the most practical ones. Expanding thescope of FPA to other regions, methods and theories will help to further consolidateits sound position in the discipline.Such applications of FPA theories will also inspire researchers in developing

countries. Small research communities in different regions can build on existingFPA theories and their nascent applications to their regions. These studies, intime, will help developing countries build their own literatures. Applying FPA the-ories to developing parts of the world may become one of the most exciting andfruitful endeavours in political science in the coming decades.

80. B. Gregory Marfleet and Colleen Miller, “Failure after 1441: Bush and Chirac in the UN SecurityCouncil”, Foreign Policy Analysis, Vol. 1, No. 3 (2005), pp. 333–360; AkanMalici, “Reagan and Gorbachev:Altercasting at the End of the Cold War”, in Schafer and Walker, op. cit., pp. 127–150; Huiyun Feng, op.cit.81. Valerie M. Hudson, “Foreign Policy Analysis: Actor-Specific Theory and the Ground of Inter-

national Relations”, Foreign Policy Analysis, Vol. 1, No. 1 (2005), pp. 1–30.82. Alexander L. George, Bridging the Gap: Theory and Practice in Foreign Policy (Washington, DC:

United States Institute of Peace, 1993).

190 Leadership Analysis at a “Great Distance”

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

Acknowledgements

I would like to thank Valerie Hudson, Klaus Brummer and anonymous reviewersfor their valuable comments and Sercan Canbolat and Toygar Halistoprak for theirresearch assistance in coding the relevant speeches.

Disclosure statement

No potential conflict of interest was reported by the author.

About the Author

Özgür Özdamar teaches at Bilkent University’s Department of InternationalRelations. He has taught at different institutions, such as the University of Missouri-Columbia, Stephens College, Ministry of Foreign Affairs of Turkey, TOBB-ETÜ, theTurkish Military Academy and the National Security College. His research focuseson foreign policy analysis, forecasting, international relations theories and securitystudies. He currently serves as editor of All Azimuth: Journal of Foreign Policy andPeace and director of research at Center for Foreign Policy and Peace Studies.

Appendix (Speeches and Tables)

Islamists – List of All Speeches Coded

Ruhollah KhomeiniKhomeini, Ruhollah, “Israel: As the enemy of Muslims and Islam”, available:<http://www.imam-khomeini.com> (accessed 28 February 2011).

Khomeini, Ruhollah, “The Granting of Capitulatory Rights to the U.S”, Islam andRevolution (Trans. Hamid Algar, Mizan Press, CA: Berkeley, 1981), pp. 181–188.

Khomeini, Ruhollah, “Islamic Government”, Islam and Revolution (Trans. HamidAlgar, Mizan Press, CA: Berkeley, 1981), pp. 27–39.

Khomeini, Ruhollah, “Address to Monsignor Bugnini, Papal Nuncio”, Islam andRevolution (Trans. Hamid Algar, Mizan Press, CA: Berkeley, 1981), pp. 278–285.

Khomeini, Ruhollah, “Thirty Million People Have Stood Up”, Islam and Revolution(Trans. Hamid Algar, Mizan Press, CA: Berkeley, 1981), pp. 321–329.

Khomeini, Ruhollah, “New Year’s Message”, Islam and Revolution (Trans. HamidAlgar, Mizan Press, CA: Berkeley, 1981), pp. 286–294.

Khomeini, Ruhollah, “Message to the Muslim Students in North America”, Islamand Revolution (Trans. Hamid Algar, Mizan Press, CA: Berkeley, 1981), pp. 209–211.

Note:All speeches are in English. They were translated from Farsi by the publisher.

Necmettin ErbakanErbakan, Necmettin, “Our Industrial Cause and Zionist Obstacles”, <http://www.necmettinerbakan.org> (accessed 28 February 2011).

Özgür Özdamar 191

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

Erbakan, Necmettin, “International Relations and Foreign Policy” (26 April 1980),<http://www.necmettinerbakan.org> (accessed 28 February 2011).

Erbakan, Necmettin, “On the Priorities of Foreign Policy with Prof. Dr NecmettinErbakan”, Interview: Hasan Hüseyin Ceylan. Dış Politika, 3rd Issue. (October 1988)

Erbakan, Necmettin, “Gulf Crises”, Ankara Mülkiyeliler Birliği, 1990.

Erbakan, Necmettin, “The Common Market and Turkey”, Elazığ, 1987, <http://www.necmettinerbakan.org> (accessed 28 February 2011).

Note: All speeches are in Turkish and original.

Muammar QaddafiQaddafi, Muammar. “Exploitation”, My [Political] Views, Trans. Mehmet Keskin,Hareket Press, Istanbul, 1974, pp. 11–56.

Qaddafi, Muammar. “The Third Universal [State] Theory”, My [Political] Views,Trans. Mehmet Keskin, Hareket Press, Istanbul, 1974, pp. 56–94.

Note: All speeches are in Turkish. They were translated from Arabic to Turkish bythe publisher.

Islamists – Codings and Positive/Negative Attributions to Self/Other (Scores)

NecmettinErbakan Speech Title

Negative Positive Sanayi Davamız ve Siyonist EngelleriSelf 27 34Other 17 6

Negative Positive Uluslararası İlişkiler Dış Politika (MeclisKürsüsünden)

S 2 17O 42 0

Negative Positive Prof.Dr.Necmettin Erbakan ile DışPolitika Öncelikleri Üzerine

S 0 9O 26 1

Negative Positive Mülkiyeliler BirliğindeS 1 26O 70 6

Negative Positive Ortak Pazar ve TürkiyeS 0 15O 76 12

192 Leadership Analysis at a “Great Distance”

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

R. Khomeini Speech Title

Negative Positive İslam ve MüslümanlarınDüşmanı: İsrail

Self 17 5Other 83 0

Negative Positive The Granting of CapitulatoryRights to the U.S

S 0 22O 34 1

Negative Positive Islamic GovernmentS 8 7O 62 1

Negative Positive Address to Monsignor Bugnini,Papal Nuncio

S 0 10O 54 5

Negative Positive Thirty Million People HaveStood Up

S 1 11O 27 4

Kaddafi Speech Title

Negative Positive Mukaddime (Sömürü)Self 11 18Other 132 7

Negative Positive Üçüncü DevletNazariyesi

S 0 47O 69 0

Neo Islamists – All Speeches Coded

Recep Tayyip ErdoğanErdogan T. Erdogan, “Erdogan’s speech in AK party group meeting” (16 October2007), available <http://www.akparti.org.tr/site/haberler/arsiv/tumu> (accessed 15November 2011).

Erdogan T. Erdogan, “Erdogan’s speech in AK party group meeting” (30 October2007), available <http://www.akparti.org.tr/site/haberler/arsiv/tumu> (accessed 15November 2011).

Erdogan T. Erdogan, “Interview with Erdogan in his US visit of 2007”. Interview:2007, available <http://www.akparti.org.tr/site/haberler/arsiv/tumu> (accessed 20November 2011).

Özgür Özdamar 193

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

Erdogan T. Erdogan, “Erdogan’s speech in Turkish Parliament” (5 November 2007),available <http://www.akparti.org.tr/site/haberler/arsiv/tumu> (accessed 15November 2011).

Erdogan T. Erdogan, “Joint Press Statement with Condoleezza Rice” (2 November2008), available <http://www.akparti.org.tr/site/haberler/arsiv/tumu> (accessed 15November 2011).

Erdogan T. Erdogan, “Interview with Erdogan in his visit to Romania in 2006”,Interview: 2006. available <http://www.akparti.org.tr/site/haberler/arsiv/tumu>(accessed 15 November 2011).

Erdogan T. Erdogan, “Erdogan’s speech in Turkish Parliament” (15 January 2008),available <http://www.akparti.org.tr/site/haberler/arsiv/tumu> (accessed 15November 2011).

Erdogan T. Erdogan, “Erdogan’s speech in AK party group meeting” (12 December2006), available <http://www.akparti.org.tr/site/haberler/arsiv/tumu> (accessed 16November 2011).

Erdogan T. Erdogan, “Erdogan’s speech in AK party group meeting” (28 December2006), available <http://www.akparti.org.tr/site/haberler/arsiv/tumu> (accessed 16November 2011).

Erdogan T. Erdogan, “Erdogan’s press conference in Turkish media”(10 October2005), available <http://www.akparti.org.tr/site/haberler/arsiv/tumu> (accessed 16November 2011).

Note: All speeches are in Turkish and original.

Mahmoud AhmedinejadAhmadinejad, Mahmoud, “61st Session of the United Nations General Assembly”(September 2006).

Ahmadinejad, Mahmoud, “63rd Session of the United Nations General Assembly”(September 2008).

Ahmadinejad, Mahmoud, “65th Session of the United Nations General Assembly”(September 2010).

Ahmadinejad, Mahmoud, “Spiegel Interview with Iranian President Ahmadine-jad” (October 2009).

Ahmadinejad, Mahmoud, “Time’s Interview with Iranian President MahmoudAhmedinejad” (September 2009).

Ahmadinejad, Mahmoud, “Transcript of the Ahmadinejad Interview by Washing-ton Post” (September 13).

Ahmadinejad, Mahmoud, “66th Session of the United Nations General Assembly”(September 2011).

Note: All speeches are in English. They were translated from Farsi.

194 Leadership Analysis at a “Great Distance”

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

Neo Islamists - Codings and Positive/Negative Attributions to Self/Other (Scores)

R. Tayyip Erdogan Speech Date/Title

Negative Positive 16 October 2007, AKPparty group meeting

Self 0 15Other 17 8

Negative Positive 30 October 2007, AKPparty group meeting

S 2 36O 22 1

Negative Positive Erdogan’s PressConference in US, 2007

S 0 13O 17 6

Negative Positive 5 November 2007 inTBMM

S 0 11O 7 3

Negative Positive Joint Conference withC. Rice

S 0 9O 8 3

Negative Positive Interview with Erdogan inRomania, 2006

S 2 11O 13 0

Negative Positive 15 January 2008 in TBMMS 0 14O 11 0

Negative Positive 12 December 2006, AKPparty group meeting

S 1 21O 24 2

Negative Positive 28 November 2006, AKPparty group meeting

S 1 17O 23 3

Negative Positive Press Conference, 10October 2005

S 2 15O 20 4

Özgür Özdamar 195

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

MahmoudAhmedinejad Speech Date/Title

Negative Positive 61st Session of the UnitedNations General AssemblySeptember 2006

Self 0 3Other 25 3

P1 =−0.79; I1 = 1; P4a = 0.09;P4b = 0.91

Negative Positive 63 rd Session of the UnitedNations General AssemblySeptember 2008

S 0 5O 37 3

P1 =−0.85; I1 = 1; P4a = 0.11;P4b = 0.89

Negative Positive 65th Session of the UnitedNations General AssemblySeptember 2010

S 0 9O 25 3

P1 =−0.79; I1 = 1; P4a = 0.24;P4b = 0.76

Negative Positive Spiegel Interview with IranianPresident AhmadinejadOctober 2009

S 4 12O 23 4

P1 =−0.72; I1 = 0.5; P4a = 0.37;P4b = 0.63

Negative Positive Time’s Interview with IranianPresident MahmoudAhmedinejad

September 2009S 2 17O 23 0

P1 =−1; I1 = 0.8; P4a = 0.45;P4b = 0.55

Negative Positive Transcript of the AhmadinejadInterview by WashingtonPost

September 13 ??S 0 10O 34 3

P1 =−0.84; I1 = 1; P4a = 0.21;P4b = 0.79

Negative Positive 66th Session of the UnitedNations General Assembly

September 2011S 1 4O 33 3

P1 =−0.84; I1 = 0.6; P4a = 0.12;P4b = 0.88

196 Leadership Analysis at a “Great Distance”

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

Mashaal, Khaled, ‘‘Our People Will Never Rest’’, Interview: Spiegel (June 2006).

Mashaal, Khaled, ‘‘Hamas’ Mesh’al Lays Out New Policy Direction’’, Interview:Jordanian Al-Sabeel newspaper (September 2010).

Mashaal, Khaled, “On Occupation and Resistance’’, Interview: Al Quassamwebsite (December 2006).

Mashaal, Khaled, “Resistance is All We Have in Gaza’’, Interview: Press TV(January 2009).

Mashaal, Khaled, “The Voice of Hamas’’, Interview: Opendemocracy (November2010).

Mashaal, Khaled, “Religion, Violence and Chances for Peace’’, Interview: KenLivingstone (from NewStatesman) (September 2009).

Mashaal, Khaled, ‘‘Khaled Meshaal Interview: Hamas Chief Weighs In on Eve ofPeace Talks’’, Interview: Sharmine Narwani (August 2010).

Note: All speeches are in English. They were translated from Arabic by thepublisher.

KhaledMeshaal Speech Date/Title

Negative Positive Spiegel Interview with KhaledMashaal – 02/06/2006

Self 1 11Other 16 0

Negative Positive Interview with Meshal by Al-Sabeelnewspaper – 25/09/2010

S 0 4O 17 1

Negative Positive InterviewwithMeshal in Al Quassamwebsite – 09/12/2006

S 2 8O 17 0

Negative Positive Press TV Interview with KhaledMeshaal – 11/01/2009

S 1 6O 37 1

Negative Positive The Voice of Hamas Interview withHamas Leader Mashall byOpendemocracy – 5/11/2010

S 1 22O 53 4

Negative Positive Hamas Leader Interview by KenLivingstone (from NewStatesman)– 17/09/2009

Continued

Özgür Özdamar 197

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017

KhaledMeshaal Speech Date/Title

S 1 13O 65 0

Negative Positive Khaled Meshaal Interview conductedby Sharmine Narwani from OxfordUniversity – 31/08/2010

S 4 22O 41 1

198 Leadership Analysis at a “Great Distance”

Dow

nloa

ded

by [

Bilk

ent U

nive

rsity

] at

22:

11 2

7 N

ovem

ber

2017